Hi,
Just wondering if I currently have a pair of 2 way bookshelf speaker using a passive crossover and if I find out the crossover point, could I then buy a Behringer SUPER-X PRO CX2310 (http://www.behringer.com/CX2310/index.cfm?lang=ENG) and set it to that crossover and it would work?
Along with 4 channels of amplification.
Is it that simple?
Just wondering if I currently have a pair of 2 way bookshelf speaker using a passive crossover and if I find out the crossover point, could I then buy a Behringer SUPER-X PRO CX2310 (http://www.behringer.com/CX2310/index.cfm?lang=ENG) and set it to that crossover and it would work?
Along with 4 channels of amplification.
Is it that simple?
I want to give you a qualified yes.
If the passive speaker is exceptionally well done, you may have more issues than the passive crossover. Also, you are removing the notch filters, BCS, etc.
Will it play, absolutely. Will it sound different, almost certainly.
Will it be better, it depends on your goals, the Speakers and other variables.
Good Luck.
Doug
yes and no. First, you need to remove the original crossover so that its no longer in the signal path. Then directly wire the speakers to the terminals, and set everything up as you had mentioned.
However that makes some assumptions that are likely not right, which is that the speaker is using a text book crossover transfer function. Because of something known as Baffle diffraction, the crossover more than likely uses some form of baffle step compensation. The biggest problem with using an active crossover like you intend is that the large majority of crossovers for less than 2-3 grand do not have the ability to modify the transfer functions beyond slope and frequency. You need the ability to change the q, poles, and also parallel multiple filters.
However that makes some assumptions that are likely not right, which is that the speaker is using a text book crossover transfer function. Because of something known as Baffle diffraction, the crossover more than likely uses some form of baffle step compensation. The biggest problem with using an active crossover like you intend is that the large majority of crossovers for less than 2-3 grand do not have the ability to modify the transfer functions beyond slope and frequency. You need the ability to change the q, poles, and also parallel multiple filters.
pjpoes is right! You have to measure what is happening! If you put the "text book crossover transfer function"-activated bookshelf speakers into a bookshelf, the chances that it works are quite good if the transfer functions of both drivers are quite smooth in the crossover region. Placing the speakers in a (filled!) bookshelf is like in-wall-mounting the speakers. There will be no baffle diffraction effects. But without measuring equipment it is quite difficult.
I'm hoping some of what Fosti was trying to say was lost in translation, but if he is saying that adding stuffing, as in polyfill to the inside of an enclosure gets rid of baffle diffraction, that is completely wrong. On a book shelf the response begins to slop off around 500-700hz, no amount of stuffing will fix that. To further illustrate my point I have made up some pictures of transfer functions and the subsequent responses on a real driver response.
this is a standard 3rd order transfer function. This is what can be done with any number of analogue or digital active crossovers. This is all that can be done, the other transfer functions I will show you can not be easily replicated with these standard active crossovers (There is a way in theory, but I've never tried it, and I highly doubt people do it).
here is the response you get with that transfer function. The dark blue is the actual driver response, and the lighter blue is the response with filter. Notice there is both some rising in the response around 500hz, and then a further much worse rising caused by a mixture of cone breakup and the natural tendency for drivers to do that.
This is the transfer function I achieved using just 4 parts, and it takes into account all of the mentioned issues. This type of transfer function can not be achieved with a normal active crossover such as the mentioned one.
Notice here that the response is much smoother and would integrate far better with a tweeter. This is why I believe that for the average consumer, the use of active filters above 500hz is problematic.
this is a standard 3rd order transfer function. This is what can be done with any number of analogue or digital active crossovers. This is all that can be done, the other transfer functions I will show you can not be easily replicated with these standard active crossovers (There is a way in theory, but I've never tried it, and I highly doubt people do it).
here is the response you get with that transfer function. The dark blue is the actual driver response, and the lighter blue is the response with filter. Notice there is both some rising in the response around 500hz, and then a further much worse rising caused by a mixture of cone breakup and the natural tendency for drivers to do that.
This is the transfer function I achieved using just 4 parts, and it takes into account all of the mentioned issues. This type of transfer function can not be achieved with a normal active crossover such as the mentioned one.
Notice here that the response is much smoother and would integrate far better with a tweeter. This is why I believe that for the average consumer, the use of active filters above 500hz is problematic.
Ah sorry, Fosti, I misread that. Yeah that would be true. Though why you would want to do that is beyond me.
This still doesn't change my point though, the issues are far greater than just bs, but response issues with the speaker. While you could use eq as well, that doesn't make sense, your better off using correct filter slopes in the first place. Something tells me the original poster has lost interest though as he hasn't posted again, maybe realized that its not such a good idea.
This still doesn't change my point though, the issues are far greater than just bs, but response issues with the speaker. While you could use eq as well, that doesn't make sense, your better off using correct filter slopes in the first place. Something tells me the original poster has lost interest though as he hasn't posted again, maybe realized that its not such a good idea.
The answer depends completely on the complexity of the passive crossover that is going to be replaced.
In most cases, a complex passive design with many components (like a dozen or more) is likely to be designed properly and exhibit optimum response. Then, the active crossover replacement should be designed to produce the same acoustical response from the drivers in order to keep the existing system performance while gaining the advantages of multi-amplification. This is something that can't just be done with a simple filter like CX2310, but a more complex one like DCX2496 can do the job with the help of its parametric EQs, the phase shifters and the delay.
On the other hand, simple generic passive crossovers are easily outperformed by generic active ones like CX2310.
In most cases, a complex passive design with many components (like a dozen or more) is likely to be designed properly and exhibit optimum response. Then, the active crossover replacement should be designed to produce the same acoustical response from the drivers in order to keep the existing system performance while gaining the advantages of multi-amplification. This is something that can't just be done with a simple filter like CX2310, but a more complex one like DCX2496 can do the job with the help of its parametric EQs, the phase shifters and the delay.
On the other hand, simple generic passive crossovers are easily outperformed by generic active ones like CX2310.
If you're going to do this at all you should consider building your own active filter - then you can directly copy the topology of the passive filter you already have, or customize it if you want to try to make an improvement (but you need to know what you're doing).
But the question is (IMO) - what will be gained, if anything, by switching from passive to active? In terms of an audible difference you may end up disappointed, unless again you can improve upon the original (see first paragraph).
But the question is (IMO) - what will be gained, if anything, by switching from passive to active? In terms of an audible difference you may end up disappointed, unless again you can improve upon the original (see first paragraph).
I partly disagree with Eva's response, as you can see from my post and pictures, I managed to flatten the response with 4 components in the woofer. Add in the tweeter and its crossover and we are talking a total of roughly 8 components. That is not a complicated crossover by any means. I think any well designed book shelf, at any price, could be built this way, and frequently is. I've measured and dismantled M&K, Infinity, and Jbl speakers, mostly less than 3-500 dollars a pair, and they used competently designed crossovers. The minimum number of crossover parts you would find in a speaker in this price class is going to be 5, in my opinion, using 2nd order crossover points and a single padding resistor. I know that some really cheap crossovers use 2 components, a capacitor on the tweeter and padding resistor, and thats, it, but I can't imagine he would be asking for advice on converting to active if that was the case. Such a design wouldn't be worth the effort, and probably wouldn't have biamp terminals, which again, if he is asking this, I'm assuming they have.
The dynaudio and Montana 1st order crossovers used something like 15-20+ components per speaker in the crossover in order to achieve the true first order slopes and linear phase response. I have always thought their advertising was misleading in that regard, trying to argue that the simpler 1st order crossover put less in the signal path. Ha, not theres, you could achieve 8th order slopes with the amount of components they had in the signal path.
The dynaudio and Montana 1st order crossovers used something like 15-20+ components per speaker in the crossover in order to achieve the true first order slopes and linear phase response. I have always thought their advertising was misleading in that regard, trying to argue that the simpler 1st order crossover put less in the signal path. Ha, not theres, you could achieve 8th order slopes with the amount of components they had in the signal path.
I'd argue that it would depend on how its being used. There is no doubt in my mind, regardless of the number of ways we crossover, that active has its advantages in certain situations. It's the consumer end of things and price point we are talking about that bothers me. Obviously in Pro Audio live sound situations where reliability is of key concern, and power is in not just the 1000's but often tens of 1000's, well then, of course an active crossover is better. Not only is it power independent, but it also offers the flexibility of making the fine adjustments needed for different venues.
However, in the case of consumer level products, where really, more than a few hundred watts is not only unusual, but unnecessary, passive crossovers don't have the same limiting factors. People always try to argue the impact they have on amplifier dampening, but really, that is a pretty useless statistic when weighed against the other issues we are dealing with here. They also argue the signal loss, which is generally on the magnitude of less than 1db in a passive crossover, typically considerably less than that. In the project I just finished the crossover has two inductors in the signal path of the midbass drivers, and caused a .1db loss at 150hz from the dcr of two 20guage inductors (After running a good 50 tests in excess of 90 dbs). In other words, here, losses are minimal, and will have no real practical impact.
The real issues that a designer must consider in designing a speaker crossover is the power response and phase response across a wide polar range, most importantly in the horizontal range. This is much more easily achieved with a passive crossover. I hear people always talk about designing their own active crossovers with the same transfer functions. That's a lot easier said than done for most of us. Making an active crossover based on someone else's schematic, even modifying the poles, q's, etc. doesn't achieve a match to the transfer functions I'm talking about. You actually need the ability to have what would essentially be multiple crossovers in parallel each having the needed slope. For instance, the first crossover point might be say 3db's per octave at 500hz, then 18db's per octave at 2200hz, giving an acoustic 4th order. Thats a major over simplification, but is simply to illustrate what it is I mean when I say, it can't be done with most readily available electronic crossovers, and still couldn't be easily achieved in one we made ourselves. Then you have other things that sometimes need to be taken care of, which can never be taken care of with a low level crossover, like LCR or CR series and parallel conjugate filters for dealing with ringing or rising inductance.
Finally, and really, this should go without saying, but you need to design all of this around actual speaker measurements in the enclosure the speaker is being used in. If this isn't done, then everything else we are talking about is unimportant. Just because the manufacturer claims they used a 2khz crossover point doesn't mean that replacing that passive with an active at 2khz will yield the same crossover point. You have to take into account the summing at the crossover point to figure what what actual crossover points need to be used, and I would wager they will not likely match the specified crossover point. You might use something like an 1800hz on the woofer and 2.2khz on the tweeter, and the middle of the summation point would look to be around 2khz (for example).
However, in the case of consumer level products, where really, more than a few hundred watts is not only unusual, but unnecessary, passive crossovers don't have the same limiting factors. People always try to argue the impact they have on amplifier dampening, but really, that is a pretty useless statistic when weighed against the other issues we are dealing with here. They also argue the signal loss, which is generally on the magnitude of less than 1db in a passive crossover, typically considerably less than that. In the project I just finished the crossover has two inductors in the signal path of the midbass drivers, and caused a .1db loss at 150hz from the dcr of two 20guage inductors (After running a good 50 tests in excess of 90 dbs). In other words, here, losses are minimal, and will have no real practical impact.
The real issues that a designer must consider in designing a speaker crossover is the power response and phase response across a wide polar range, most importantly in the horizontal range. This is much more easily achieved with a passive crossover. I hear people always talk about designing their own active crossovers with the same transfer functions. That's a lot easier said than done for most of us. Making an active crossover based on someone else's schematic, even modifying the poles, q's, etc. doesn't achieve a match to the transfer functions I'm talking about. You actually need the ability to have what would essentially be multiple crossovers in parallel each having the needed slope. For instance, the first crossover point might be say 3db's per octave at 500hz, then 18db's per octave at 2200hz, giving an acoustic 4th order. Thats a major over simplification, but is simply to illustrate what it is I mean when I say, it can't be done with most readily available electronic crossovers, and still couldn't be easily achieved in one we made ourselves. Then you have other things that sometimes need to be taken care of, which can never be taken care of with a low level crossover, like LCR or CR series and parallel conjugate filters for dealing with ringing or rising inductance.
Finally, and really, this should go without saying, but you need to design all of this around actual speaker measurements in the enclosure the speaker is being used in. If this isn't done, then everything else we are talking about is unimportant. Just because the manufacturer claims they used a 2khz crossover point doesn't mean that replacing that passive with an active at 2khz will yield the same crossover point. You have to take into account the summing at the crossover point to figure what what actual crossover points need to be used, and I would wager they will not likely match the specified crossover point. You might use something like an 1800hz on the woofer and 2.2khz on the tweeter, and the middle of the summation point would look to be around 2khz (for example).
soundengine355 said:
All great replies. The simple answer is yes! If it were my project:
I would set the crossover to 24dB/octave or higher and choose
a crossover point of 6 kHz TO START. Then I would listen and by
my listening tastes raise or lower it. . .this is DIY, let's have some fun!
Make sure the two amplifiers have the same gain. . .it will make
trimming easier. I would leave any L-pads on a tweater. I would
also add a large value capacitor to the tweeter to give it protection
from any thump from amps turning on or off. If you can easily replace
the tweeter then forgo the cap. . .
You will need less power than you think when the passive crossover is
removed. . .everything will open up. There are a thousand problems
you will run into; add to the post when you run into them. . . your
ear will tell you much. . .I like it. . .I hate it. . . my wife loves it so much
that I will get lucky tonight. . .etc.
A three way x-over might be better because it would allow you
to set a different high and low x-over. . . and in the future when
you do the same to a three way speaker you will be ready and
have experience. . .
You will be happy you made the change. . .
I'm pretty much with Cris above. Depending on your goals. Just want to have fun and learn a few things? - yeah - go right ahead. You might like what you hear. 
But if you have serious ideas about making your bookshelf speakers into something magically wonderful, then you should probably save your money.
That crossover is very limited - you only get 24dB/octave filters and you can't stagger the frequencies nor adjust phase. But you might not have to. And you don't have to cross at the same place the passive does.
It does have a sub channel, too - if ever you wanted it.
I'd say go into it knowing that "yes, it will work" but it may not be better. But you'll learn a lot doing it!
But if you have serious ideas about making your bookshelf speakers into something magically wonderful, then you should probably save your money.
That crossover is very limited - you only get 24dB/octave filters and you can't stagger the frequencies nor adjust phase. But you might not have to. And you don't have to cross at the same place the passive does.
It does have a sub channel, too - if ever you wanted it.
I'd say go into it knowing that "yes, it will work" but it may not be better. But you'll learn a lot doing it!
Passive crossovers have many limitations because they rely completely on driver impedance curves assuming that they can't change, but these curves change when anything above a few watts is applied to the drivers. Impedance does even become non-linear when some diaphragm displacement is involved and when voice coil inductance dominates due to hysteresis. Furthermore, parameters such as midrange and tweeter resonant frequencies and Qms change with aging resulting in detuned passive crossovers.
With an active crossover, system performance is no longer dependent on driver impedance.
Passive crossovers with low part counts only produce good performance with very well behaved drivers, which are quite unusual, or at least hardly affordable.
With an active crossover, system performance is no longer dependent on driver impedance.
Passive crossovers with low part counts only produce good performance with very well behaved drivers, which are quite unusual, or at least hardly affordable.
Re: Re: Changing a passive crossover to active
Hi,
The simple answer is is no. THe complicated answer is no.
It is a complete waste of time. The only cases where it might
not make things worse, where the crossover is wrong in the
firstplace, are not worth bothering with and they will still be
wrong with a standard function active crossover.
Active 2-ways are a waste of exspense driving the tweeters
unless you have suitably lower power amplifiers for them.
The Beheringer c/o is 4th order L/R electrical variable.
It has nothing to do with crossover transfer functions.
It would ruin any decent 2-way design.
/sreten.
gni said:
Hi,
The simple answer is is no. THe complicated answer is no.
It is a complete waste of time. The only cases where it might
not make things worse, where the crossover is wrong in the
firstplace, are not worth bothering with and they will still be
wrong with a standard function active crossover.
Active 2-ways are a waste of exspense driving the tweeters
unless you have suitably lower power amplifiers for them.
The Beheringer c/o is 4th order L/R electrical variable.
It has nothing to do with crossover transfer functions.
It would ruin any decent 2-way design.
/sreten.
soundengine355,
- planning to do the same, experimenting with same Behringer CX2310, LM3886 chipamps and Beovox MC120.2 three-ways.
- a few "L/R, active crossover" articles at Elliot Sound on the topic:
Benefits of Bi-Amplification - Not quite magic, but close (Part I)
http://sound.westhost.com/bi-amp.htm
Benefits of Bi-Amplification - Crossovers, tri-amping, etc. (Part II)
http://sound.westhost.com/bi-amp2.htm
Active Vs. Passive Crossovers
http://sound.westhost.com/biamp-vs-passive.htm
Baffle Step Compensation
http://sound.westhost.com/bafflestep.htm
Linkwitz Transform Circuit
http://sound.westhost.com/linkwitz-transform.htm
- planning to do the same, experimenting with same Behringer CX2310, LM3886 chipamps and Beovox MC120.2 three-ways.
- a few "L/R, active crossover" articles at Elliot Sound on the topic:
Benefits of Bi-Amplification - Not quite magic, but close (Part I)
http://sound.westhost.com/bi-amp.htm
Benefits of Bi-Amplification - Crossovers, tri-amping, etc. (Part II)
http://sound.westhost.com/bi-amp2.htm
Active Vs. Passive Crossovers
http://sound.westhost.com/biamp-vs-passive.htm
Baffle Step Compensation
http://sound.westhost.com/bafflestep.htm
Linkwitz Transform Circuit
http://sound.westhost.com/linkwitz-transform.htm
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.